Skip to contents

Compute Grouped Numeric Summary Statistic

Source: R/pacc.R
compute_score_summary.Rd

This function is used to compute the numeric variable summary statistic grouped by categorical variable.

Usage

compute_score_summary(
  .data,
  wideFormat = TRUE,
  scoreVar,
  groupVar = "DX",
  filterGroup = NULL,
  groupVar1 = NULL
)

Arguments

.data

Data.frame

wideFormat

A Boolean value whether the input data.frame is in a wide or long format, Default: TRUE

scoreVar

Character vector of variable(s) that contain the actual score/numeric values

When wideFormat is FALSE (i.e., for a long format data), scoreVar must be a length of one character vector of variable name that contains the score/numeric values.

groupVar

Group variable, Default: 'DX'

filterGroup

Filter value of group variable groupVar, Default: NULL Only applicable if the groupVar is a length of one character vector.

groupVar1

Additional grouping variable, only applicable for long format data.frame.

Value

A data.frame with the following columns:

  • groupVar: Grouping variable

  • VAR: Score/numeric variable name

  • N: Number of non-missing observation

  • MEAN: Mean score

  • SD: Standard deviation value

Details

All computed summary statistic are based on non-missing observation. The result summary will be filter by the corresponding filterGroup value(s).

See also

compute_baseline_score_summary() vignette(topic = "ADNIMERGE2-PACC-SCORE", package = "ADNIMERGE2")

Examples

if (FALSE) { # \dontrun{
# For long format data
# Suppose we wanted to compute the baseline summary score of
# all available assessments in \code{ADNIMERGE2::ADQS}

# By baseline diagnosis status.
library(tidyverse)
library(ADNIMERGE2)

long_format_example <- ADNIMERGE2::ADQS %>%
  filter(ABLFL %in% "Y") %>%
  #  Check there is only one baseline record per assessment type per subject
  ADNIMERGE2::assert_uniq(USUBJID, PARAMCD)

compute_score_summary(
  .data = long_format_example,
  wideFormat = FALSE,
  scoreVar = "AVAL",
  groupVar1 = "PARAMCD",
  groupVar = "DX",
  filterGroup = NULL
)

# For only cognitive normal (CN) subjects
compute_score_summary(
  .data = long_format_example,
  wideFormat = FALSE,
  scoreVar = "AVAL",
  groupVar1 = "PARAMCD",
  groupVar = "DX",
  filterGroup = "CN"
)

# For wide format data
# Suppose we wanted to compute the baseline summary statistic of
# \code{AGE}, \code{BMI}, \code{ADASTT11} and \code{ADASTT13}
wide_format_example <- ADNIMERGE2::ADSL %>%
  filter(ENRLFL %in% "Y")

# By baseline diagnostics status
compute_score_summary(
  .data = wide_format_example,
  wideFormat = TRUE,
  scoreVar = c("AGE", "BMI", "ADASTT11", "ADASTT13"),
  groupVar1 = NULL,
  groupVar = "DX",
  filterGroup = NULL
)


compute_score_summary(
  .data = wide_format_example,
  wideFormat = TRUE,
  scoreVar = c("AGE", "BMI", "ADASTT11", "ADASTT13"),
  groupVar1 = NULL,
  groupVar = "DX",
  filterGroup = "CN"
)

# By SEX
compute_score_summary(
  .data = wide_format_example,
  wideFormat = TRUE,
  scoreVar = c("AGE", "BMI", "ADASTT11", "ADASTT13"),
  groupVar1 = NULL,
  groupVar = "SEX",
  filterGroup = NULL
)
} # }